Mar Biol DOI 10.1007/s00227-007-0717-x

RESEARCH ARTICLE

Substratum preferences in planula larvae of two of scleractinian , Goniastrea retiformis and Stylaraea punctata

Yimnang Golbuu · Robert H. Richmond

Received: 10 July 2006 / Accepted: 23 April 2007 © Springer-Verlag 2007

Abstract To test whether planulae recruit randomly that the preferred substrata among those tested bears a relation- to diVerent habitats or have speciWc substratum ship to the habitats in which adult colonies were found. preferences, the settling behavior of planulae from two shallow water coral species from Pago Bay, Guam (13°25.02N, 144°47.30E) were examined in the laboratory in June and Introduction July of 1995. Goniastrea retiformis is generally restricted to the shallow reef front (<10 m depth) in areas dominated Many sessile, benthic marine invertebrates have planktonic by crustose (CCA), while Stylaraea punctata larvae that diVer morphologically and physiologically from is abundant on inner reef Xats were CCA coverage is low the juvenile and adult forms, and constitute the dispersal and sand and carbonate rubble covered by bioWlms is com- stage in the organisms’ life history. The period during mon. When presented with four substrata (1) carbonate which these larvae can successfully settle and metamor- rock scrubbed free of bioWlm and dried as a control, (2) the phose into the juvenile form is called the competency CCA Hydrolithon reinboldii, (3) the CCA Peyssonelia sp., period, and is aVected by both intrinsic and extrinsic factors and (4) naturally conditioned carbonate rubble covered by a including nutritional mode, resource availability, tempera- bioWlm, G. retiformis larvae showed a signiWcant preference ture and the presence of speciWc chemical cues (Richmond for H. reinboldii, and S. punctata larvae for the carbonate 1987). Successful reproduction is only the Wrst step in the bioWlm treatment. The preference shown by S. punctata process of population replenishment, maintenance and larvae for bioWlmed surfaces did not diminish with increasing growth. Larvae must recruit, that is, settle and metamor- larval age up to 11 days. These results suggest that the phose, in order to become part of the population. Settlement larvae of both species are capable of habitat selection, and is a physical process, during which larvae leave the water column and come into contact with the substratum, and is often reversible, with larvae contacting unsuitable substrata returning to the water column. Metamorphosis is a physio- logical process, during which morphological, physiological Communicated by J. P. Grassle. and metabolic changes occur, and with few exceptions, is Electronic supplementary material The online version of this non-reversible (Richmond 1985). Once larvae settle and article (doi:10.1007/s00227-007-0717-x) contains supplementary come into contact with the substratum, chemicals associ- material, which is available to authorized users. ated with the substratum, conspeciWcs or preferred prey may be responsible for metamorphic induction (Morse Y. Golbuu (&) Palau International Coral Reef Center, 1990; Pawlik and HadWeld 1990; HadWeld and Paul 2001). P.O. Box 7086, Koror, PW 96940, Palau Studies of larval recruitment in corals can help determine e-mail: [email protected] the extent to which distribution patterns are shaped by post- metamorphic events, with settlement and metamorphosis R. H. Richmond Kewalo Marine Laboratory, University of Hawaii at Manoa, occurring randomly in available space, rather than pre- 41 Ahui Street, Honolulu, HI 96813, USA metamorphic events, such as the organisms’ responses to 123 Mar Biol speciWc environmental factors (i.e., speciWc sites and chem- and reproductive mode. On Guam, it is mainly restricted ical cues), in shaping the Wnal observed distribution to the shallow forereef (<10 m), an area dominated by patterns (Babcock and Mundy 1996; Morse et al. 1988; crustose coralline algae. G. retiformis is a simultaneous Mundy and Babcock 2000; Carlon 2002; Baird et al. 2003). hermaphroditic spawning species (Richmond and Hunter Several studies have examined settlement preferences of 1990). During spawning, eggs and sperm are packaged coral planula larvae. Planulae of some corals do not exhibit together within individual polyps, and released in bundles. any apparent speciWcity. For example, Pocillopora dami- The eggs are buoyant and rise to the surface of the water, cornis larvae can settle and metamorphose on almost any carrying the sperm with them. After the egg and sperm bun- hard surface as long as it is covered with biological Wlms dles reach the surface of the water, the gametes separate, (Harrigan 1972) and these larvae are relatively unaVected and fertilization takes place. Eggs from G. retiformis did by reductions in light underneath hyacinthus not contain zooxanthellae from their maternal line, but colonies (Baird and Hughes 2000). Similarly, Stylophora rather, acquired them from the environment. The mean pistillata larvae metamorphosed in all treatment assays oocyte diameter for G. retiformis is about 300 m with a examined, including unWltered seawater (Baird and Morse range of 270–340 m. 2004). Larvae of the corals Porites porites (Goreau et al. In contrast, Stylaraea punctata is only found on the shal- 1981) and Favia fragum (Lewis 1974) also seem to have low water areas of inner reef Xats on Guam, and reproduces little substratum speciWcity. by brooding larvae. It is the smallest of all zooxanthellate In contrast, some corals are very speciWc in terms of scleractinian corals, usually growing to no more than where they settle and metamorphose. Morse et al. (1988) 20 mm in diameter. The larvae of S. punctata show plastic- found that humilis larvae are very spe- ity in form and change from an elongated form to a pear- ciWc, diVerentiating among crustose coralline algae, and shaped form while swimming and crawling. The elongated only settling and metamorphosing on certain CCA species. form measured ca. 600 £ 400 m–900 £ 300 m. The and A. agaricites danai larvae also pear-shaped larvae were ca. 500 £ 400 m in size. show settlement preferences for coralline algae, but the S. punctata was of particular interest because crustose requirement for speciWc coralline algae is not as stringent as coralline algae were observed to overgrow colonies in the in A. agaricites humilis. Acropora tenuis and A. millepora Weld (Fig. 1). While many corals studied so far appear to larvae have also been found to preferentially settle on the prefer substrata covered with coralline algae, S. punctata CCA Titanoderma prototypum, over other CCA species did not follow that pattern. In the Weld, colonies of S. punc- (Harrington et al. 2004). Acropora nasuta, A. tenuis and tata are found in areas where crustose coralline algae cover A. digitifera also require speciWc CCA to settle and meta- is low, while sand and bare substrata covered with micro- morphose (Morse et al. 1996). bial/diatomaceous Wlms are common. Crustose coralline red algae (CCA) induce settlement In this paper, we present the results of experiments per- and metamorphosis in many corals (Morse et al. 1988; formed to address larval substratum preferences in these Heyward and Negri 1999; Raimondi and Morse 2000; two species of corals, and how this may contribute to the Baird and Morse 2004; Harrington et al. 2004). Crustose observed distribution patterns of these species. We hypoth- coralline algae dominate reef front areas, so coral larvae esize that larval selectivity is related to habitat distribution. might be using them as indicators of suitable habitats and conditions. The actual chemical that triggers metamorpho- sis in corals can be found in the cell walls of the CCA (Morse and Morse 1991) or it can be from bacteria associ- ated with the alga’s surface (Negri et al. 2001). Larvae of the coral Acropora willisae and A. millepora were induced to metamorphose by Pseudoalteromonas bacteria isolated from the coralline algae Hydrolithon onkodes (Negri et al. 2001). In addition, Baird and Morse (2004) suggested that both large and small water borne molecules induce meta- morphosis in Stylophora pistillata planulae, many of which may be of bacterial origin. To determine the role of larval selectivity in aVecting coral recruitment patterns, two species were selected for study in Guam, Micronesia that exhibit diVerent reproduc- tive and distribution characteristics. The coral Goniastrea Fig. 1 Crustose red coralline alga Hydrolithon reinboldii overgrow- retiformis was selected for its habitat distribution pattern ing a colony of Stylaraea punctata. Scale bar 1mm 123 Mar Biol

To test this hypothesis, we compared the larval selectivity Jokiel (1984) for Pocillopora damicornis, another brooding of G. retiformis and S. punctata larvae. species. Corals were placed into 3 l containers that over- Xowed into larval collectors with sides made from 45-m nylon screens. Larvae were collected every day. Collected Materials and methods larvae were maintained in UV sterilized and Millipore Wltered seawater until they were ready to be used in the Collection of coral colonies experiments. The adult colonies and larvae of S. punctata were maintained outside under natural light and ambient Colonies of G. retiformis and S. punctata were collected water temperatures of 27–28°C. from Pago Bay, Guam, (13°25.02N, 144°47.30E) in 1995 for this study. G. retiformis were sampled in the Weld 1– Substratum preferences 2 weeks before the expected spawning time to determine whether they were reproductively mature (Richmond and Goniastrea retiformis larvae Hunter 1990). The presence of colored eggs indicated that the colonies were ripe. Ripe colonies were collected a few Experiments were carried out in 250-ml glass jars contain- days before the June full moon and maintained in Xow ing 150-ml of UV sterilized and Millipore Wltered seawater through seawater tanks. Commencing the night of the June and one piece of test substratum. The following substrata full moon, colonies were observed each night for spawning were tested: pieces of the crustose coralline alga, Hydroli- activity and to collect gametes for controlled crosses. thon reinboldii, the crustose red alga, Peyssonelia sp. and Colonies of S. punctata were collected from the reef Xat pieces of naturally conditioned carbonate rubble on which at Pago Bay, Guam. Only colonies larger than 10 mm no coralline algae were growing. The naturally conditioned in diameter were collected since smaller colonies may not carbonate rubble pieces were collected from the Weld and have been reproductively mature. Coral colonies were placed in the water tables until they were used in the exper- placed in 3-l plastic containers under running seawater. iments. Reef rocks that were scrubbed and dried to get rid of coralline algae or microbial Wlms were used as controls. Collection of larvae Six replicate jars for each treatment were set up simulta- neously with a single piece of test substratum in each, and For G. retiformis, the procedures used for fertilization of 15, 3-day-old larvae were placed into each. Experimental eggs and rearing of larvae were as described by Richmond jars were arranged in a randomized block design. We did (1988). During spawning, coral gamete bundles were col- not aerate the experimental jars; preliminary work showed lected and the eggs and sperm separated using 45-m nylon that we got better recruitment rates if we did not bubble air screens. The sperm were small enough to pass through the into the vessels. 45-m nylon screens while the eggs were retained. Sperm were washed from the eggs using UV sterilized and Milli- Stylaraea punctata larvae pore Wltered seawater and were collected in a beaker. Fertil- izations were performed by adding 1 ml of a diluted sperm The experiments on the S. punctata larvae were set up in a suspension obtained from one coral colony, to a glass bea- diVerent way, with the same four substrata treatments being ker containing about 300 eggs from a diVerent colony, placed together, but not touching one another, on the Xoor of resulting in a sperm concentration of about 105 sperm ml¡1. each of the six replicate containers. Due to the limited num- This concentration ensures suYcient sperm to fertilize the ber of competent S. punctata larvae, only ten larvae were eggs but was found to be low enough to prevent poly- added to each container and the larvae were <1 day old. spermy. Several coral colonies were used to make several crosses to yield enough larvae to conduct the experiments. EVect of time on substratum preference of Stylaraea Outcrossing was found to yield fertilization rates exceeding punctata larvae 90%, and high quality larvae. After fertilization, developing embryos were kept in UV sterilized seawater at a density of To study the eVect of time following fertilization on substra- ca. 1 larva ml¡1. The embryos and larvae were cultured tum preference and selectivity, the experiment described outside under natural light and at ambient temperatures above was repeated with 3-, 7-, 9- and 11-day-old larvae. of 27–28°C. We observed Goniastrea larvae to be fully ciliated and mobile within 18 h with the Wrst substratum Statistical analysis exploration and settlement behavior observed within 72 h. The collection of S. punctata larvae from the parent col- The experiment was scored after 24 h by counting the num- onies followed the procedures described by Richmond and ber of metamorphosed individuals and the total number of 123 Mar Biol individuals recovered. Data were expressed as the propor- 70 tion of larvae that had metamorphosed. The total number 60 of individuals for calculations included free-swimming, settled and metamorphosed larvae. A Kruskal–Wallis 50 ANOVA was used to test for diVerences among the diVer- 40 ent substrata for the preference experiments, while one-way 30 ANOVA was used to assess the diVerence in preference of S. punctata larvae over time. 20 % Metamorphosed 10

0 Results and discussion NCCR Control

Preference of Goniastrea retiformis larvae H. reinboldii Peyssonelia sp. Treatment There were signiWcant diVerences in the number of larvae that metamorphosed on diVerent substrata (P =0.006; Fig. 3 Substratum preference of 1-day-old Stylaraea punctata larvae. Fig. 2). More larvae settled on the crustose coralline alga, There were six replicates of each treatment with ten larvae per repli- cate. NCCR naturally conditioned carbonate rubble. Kruskal–Wallis H. reinboldii than on the other substrata (Fig. 2). There was analysis revealed signiWcant diVerences among treatments (P = 0.001). no settlement or metamorphosis on the controls (Fig. 2). At Values shown are mean § SE the end of the experiment, most of the larvae had attached to the substratum, but had not fully metamorphosed. The larvae that had attached and formed a basal plate were EVect of time on preference of Stylaraea punctata larvae counted as metamorphosed. The preference of S. punctata larvae for the microbial Wlm Preference of Stylaraea punctata larvae treatment, H. reinboldii, Peyssonelia sp., and the control did not signiWcantly change over a period of eleven days There were signiWcant diVerences in the number of larvae (Table 1). that settled and metamorphosed on the diVerent substrata The average recovery (free-swimming, settled and meta- (P = 0.001; Fig. 3). SigniWcantly more larvae settled and morphosed) for the G. retiformis preference experiment metamorphosed on the microbial Wlm than on the other sub- was 79% while the recovery for the <1 day old S. punctata strata (P <0.05, n =6; Fig.3). No larvae settled and meta- preference experiment was 73% (Supplementary table). morphosed on the control substrata (Fig. 3). The recovery rate was not related to the age of the larvae used. The S. punctata preference experiment using 9-day- old larvae had the highest recovery at 83% while 7-day-old 70 larvae had the lowest at 65% (Supplementary table). Larval

60 losses were probably the result of mortality and decomposi- tion since we did not see any recruitment on the walls of the 50 experimental containers. 40 The results of this research demonstrated two distinctly V 30 di erent recruitment responses in coral larvae from two diVerent species of stony corals: a preference for the crus- 20 tose coralline alga H. reinboldii in Gonistrea retiformis, % Metamorphosed 10 and an avoidance of this substratum, with a preference for W 0 bio lms in S. punctata. Previous studies have demonstrated that some corals do not have strict preferences for a particu- NCCR Control lar substratum during settlement and metamorphosis (Baird

H. reinboldii and Morse 2004; Goreau et al. 1981; Harrigan 1972; Lewis Peyssonelia sp. Treatment 1974), while other corals only settle and metamorphose on speciWc substrata (Morse et al. 1988, 1996; Morse and Fig. 2 Substratum preference of 3-day-old Goniastrea retiformis lar- Morse 1991; Negri et al. 2001; Raimondi and Morse 2000; vae. There were six replicates of each treatment with 15 larvae per rep- licate. NCCR naturally conditioned carbonate rubble. Kruskal–Wallis Baird and Morse 2004; Harrington et al. 2004). Because ANOVA revealed signiWcant diVerences among the treatments S. punctata reproduce by brooding, we expected it to be (P = 0.006). Values shown are mean § SE non-selective like other brooders that have been studied before 123 Mar Biol

Table 1 Settlement preference of S. punctata larvae over time istic of some crustose coralline algal species is the regular shedding of their surface layers of cells, which may reduce Time Control H. reinboldii Peyssonellia Naturally (days) sp. conditioned epiphytic and epizootic competitors for space. Most corals carbonate rubble readily overgrow coralline algae, thus this substratum is also appropriate from a space and competition perspective. 10 2.30 58.5 The preference of S. punctata larvae for microbial Wlm 3 0 1.7 1.7 53.0 did not decrease over time. Studies of several diVerent 70 2.80 65.1 invertebrate species have shown that larvae may lose selec- 90 0 0 64.5 tivity when the larval phase is prolonged (Coon et al.1990; 11 0 5.5 0 44.2 Rittschof et al. 1984; Highsmith 1982). The maintenance of Values shown are percent of larvae that metamorphosed on each sub- stringency and speciWcity in S. punctata demonstrated in strata. No signiWcant diVerence in preference for microbial Wlms over this study is similar to what Morse et al. (1996) demon- the 5 days tested (one-way ANOVA, P = 0.34) strated for larvae of several Acropora species. The reason why preference in S. punctata larvae did not decrease over (Harrigan 1972; Lewis 1974; Goreau et al. 1981; Baird and time might be attributed to the length of the experiments. Morse 2004). The results of this work show that larvae of S. The experiment was designed so that all of the choices were punctata are highly speciWc in where they settle and meta- available to the larvae. If only one substratum was pre- morphose, in both preference and apparently avoidance sented to the larvae, then the larvae might have lost selec- (Fig. 3, Table 1). The results indicate that reproductive tivity and settled and metamorphosed on that substratum. mode is not a good predictor of whether coral larvae will be It is also possible that since all the choices were avail- speciWc in where they settle and metamorphose, rather, able to S. punctata larvae, they could have received the adult distribution is a better indicator of whether larvae will metamorphic inducer from one substratum and metamor- be speciWc or general with respect to recruitment patterns. phosed on another. But if this were the case, we would S. punctata colonies have a limited distribution pattern that expect the number of metamorphosed larvae on all of the is similar to that of spawners and unlike many brooders that test substrata to be similar. The results did not show this have widely distributed adult colonies. outcome, rather it showed that most larvae metamorphosed Adult colonies of S. punctata were frequently found over- on naturally conditioned surfaces (Fig. 3, Table 1). grown by coralline algae in the Weld (Fig. 1) as a result of The more likely reason why preference does not their small size and slow growth rate. G. retiformis recruits decrease through time in S. punctata larvae in contrast with are only susceptible to being overgrown by coralline algae some other invertebrate larvae may be because the mainte- when they are juveniles. Once they reach a certain size, cor- nance of larval substrata preference is critical for survival. alline algae are not able to overgrow them. Since S. punctata Individuals that recruit to carbonate rubble within the shal- have small colonies, they never attain a refuge in size where low water reef Xat habitat in which they are found have a they cannot be overgrown by coralline algae. Thus, the survi- better chance of survival than those that lose their selectiv- vorship of larvae avoiding crustose coralline algae is ity over time and settle on coralline algae, or disperse to expected to be higher than for those larvae that recruit on other habitats more favorable to other species, since recruits faster growing species of crustose coralline algae. will have a greater likelihood of being overgrown. While the speciWc substratum-associated metamorphic Two models have been used to describe recruitment pat- inducers addressed in this study may have diVered, both terns of marine invertebrates (Morse et al. 1988). In the types of larval responses support habitat selection and “lottery” model, recruitment occurs randomly when space enhanced levels of post-metamorphic survivorship. The becomes available while in the “deterministic” model, the preference of G. retiformis larvae for crustose coralline larval selectivity for appropriate substrata is important in algae (Figs. 2, 3) is similar to that observed in several other determining spatial patterns in recruitment (Morse et al. coral species studied (Morse et al. 1988, 1996; Morse and 1988). The results of this study support the “deterministic” Morse 1991; Heyward and Negri 1999; Raimondi and model. Morse 2000; Negri et al. 2001; Baird and Morse 2004; An interesting question that still needs to be resolved is Harrington et al. 2004). Crustose coralline algae may act as the nature and source of the metamorphic inducers for an environmental indicator of adequate light, water motion G. retiformis and S. punctata. Previous studies have raised and water quality. Additionally, juvenile corals can be questions as to whether inducers associated with crustose smothered by turf and Xeshy alga and avoidance of these coralline algae and other bioWlms are of algal or bacterial types of substrata is adaptive in corals. origin. Alternatively, there may be chemicals produced by Substrata covered with crustose coralline algae do not coralline algae and other organisms that serve to inhibit have high coverage of turf and Xeshy algae. One character- recruitment of S. punctata and other types of larvae. 123 Mar Biol

Additional research on recruitment speciWcity is needed Harrigan JS (1972) The planulae larvae of Pocillopora damicornis, on the growing number of corals from which planulae lunar periodicity of swarming and substratum selection behavior. Ph.D. thesis, University of Hawaii, Hawaii, p 319 larvae can be collected or raised. The role of competition, Harrington L, Fabricius K, De’ath G, Negri A (2004) Recognition and predation, and disturbance in shaping distribution patterns selection of settlement substrata determine post-settlement sur- have been well documented, but there are relatively few vival in corals. Ecology 85:3428–3437 studies on the role of larval habitat selection in setting dis- Heyward AJ, Negri AP (1999) Natural inducers for coral larval meta- W morphosis. Coral Reefs 18:273–279 tribution patterns (reviewed by Pawlik and Had eld 1990). Highsmith RC (1982) Induced settlement and metamorphosis of sand Additional research testing particular substrata that induce dollar (Dendraster excentricus) larvae in predator-free sites: adult settlement and metamorphosis in corals will contribute to sand dollar beds. Ecology 63:329–337 the knowledge of important factors that shape distribution Lewis JB (1974) The settlement behavior of planulae larvae of the her- matypic coral Favia fragum (Esper). J Exp Mar Biol Ecol 15:165– patterns and are responsible for the persistence of coral 172 reefs. Likewise, studies of recruitment inhibition are Morse DE (1990) Recent progress in larval settlement and metamor- needed to better understand coral reef ecosystem dynamics. phosis: closing the gaps between molecular biology and ecology. Bull Mar Sci 46:465–483 Acknowledgments Research support was provided by the NIH- Morse DE, Morse ANC (1991) Enzymatic characterization of the MBRS, EPA-STAR and NOAA CSCOR/CRES programs. The authors morphogen recognized by Agaricia humilis (scleractinian coral) are grateful to Gustav Paulay for his guidance and helpful discussions larvae. Biol Bull 181:104–122 regarding this work. We thank three anonymous reviewers for helpful Morse DE, Hooker N, Morse ANC, Jensen RA (1988) Control of larval comments on the manuscript. All of the experiments conducted com- metamorphosis and recruitment in sympatric agariciid corals. plied with the current laws of the country in which they were per- J Exp Mar Biol Ecol 116:193–217 formed. Morse ANC, Iwao K, Baba M, Shimoike K, Hayashibara T, Omori M (1996) An ancient chemosensory mechanism brings new life to coral reefs. Biol Bull 191:149–154 Mundy C, Babcock R (2000) Are vertical distribution patterns of scle- References ractinian corals maintained by pre- or post-settlement processes? A case study of three contrasting species. Mar Ecol Prog Ser Babcock R, Mundy C (1996) Coral recruitment: consequences of set- 198:109–119 tlement choice for early growth and survivorship in two sclerac- Negri AP, Webster NS, Hill RT, Heyward AJ (2001) Metamorphosis tinians. J Exp Mar Biol Ecol 206:179–201 of broadcast spawning corals in response to bacteria isolated from Baird AH, Hughes TP (2000) Competitive dominance by tabular crustose algae. Mar Ecol Prog Ser 223:121–131 corals: an experimental analysis of recruitment and survival of Pawlik JR, HadWeld MG (1990) A symposium on chemical factors that understorey assemblages. J Exp Mar Biol Ecol 251:117–132 inXuence the settlement and metamorphosis of marine invertebrate Baird AH, Morse ANC (2004) Induction of metamorphosis in larvae larvae: introduction and perspective. Bull Mar Sci 46:450–454 of the brooding corals Acropora palifera and Stylophora pistilla- Raimondi PT, Morse ANC (2000) The consequences of complex larval ta. Mar Freshw Res 55:469–472 behavior in a coral. Ecology 81:3193–3211 Baird AH, Babcock RC, Mundy CP (2003) Habitat selection by larvae Richmond RH (1985) Reversible metamorphosis in coral planulae lar- inXuences the depth distribution of six common coral species. vae. Mar Ecol Prog Ser 22:181–185 Mar Ecol Prog Ser 252:289–293 Richmond RH (1987) Energetics, competency, and long-distance Carlon DB (2002) Production and supply of larvae as determinants of dispersal of planula larvae of the coral Pocillopora damicornis. zonation in a brooding tropical coral. J Exp Mar Biol Ecol Mar Biol 93:527–533 268:33–46 Richmond RH (1988) Competency and dispersal potential of planula Coon SL, Fitt WK, Bonar DB (1990) Competence and delay of meta- larvae of a spawning versus a brooding coral. Proc 6th Int Coral morphosis in the PaciWc oyster Crassostrea gigas. Mar Biol Reef Symp 2:827–831 106:379–387 Richmond RH, Hunter CC (1990) Reproduction and recruitment of Goreau NI, Goreau TJ, Hayes AL (1981) Settling, survivorship and corals: comparisons among Caribbean, the Tropical PaciWc, and spatial aggregation in planulae and juveniles of the coral Porites the Red Sea. Mar Ecol Prog Ser 60:185–203 porites (Pallas). Bull Mar Sci 31:424–435 Richmond RH, Jokiel PL (1984) Lunar periodicity in larva release in HadWeld MG, Paul VJ (2001) Natural chemical cues for settlement and the reef coral Pocillopora damicornis. Mar Biol 93:527–533 metamorphosis of marine invertebrate larvae. In: McClintock JB, Rittschof D, Branscomb ES, Costlow JD (1984) Settlement and behav- Baker BJ (eds) Marine chemical ecology. CRC Press, Boca ior in relation to Xow and surface in larval barnacles, Balanus Raton, pp 431–462 amphitrite Darwin. J Exp Mar Biol Ecol 82:31–146

123